---
datasets:
- bentrevett/multi30k
language:
- en
metrics:
- accuracy
pipeline_tag: feature-extraction
---
# mCLIP: multilingual CLIP

Official checkpoints for ACL 2023 long paper "mCLIP: Multilingual CLIP via Cross-lingual Transfer" [[paper](https://aclanthology.org/2023.acl-long.728/)]

The models are expected to be loaded with [fairseq v0.10.2](https://github.com/facebookresearch/fairseq/tree/v0.10.2) toolkit. Currently, we do not have the additional resources to migrate this project to Huggingface ecosystem. We are appreciate if you would like to contribute to the migration. 

More details are available at <https://github.com/ghchen18/acl23_mclip>.